Inorganic Chemistry
1082 NOTES
COKTRIBUTION FROM THE CHEMISTRY DEPARTMEST, Stark receiver. The reaction mixture was concentrated by evapREACTION MOTORS DIVISION, THIOKOL CHEMICAL orating most of the solvent and then dissolved in methanol and precipitated by addition of water. The solid precipitate weighed CORPORATION, DENVILLE, NEW JERSEY 19.5 g. (407, yield); a sample twice recrystallized from hexane Infrared analysis, molecular weight demelted a t 249-251 termination, and elemental analysis indicated the compound to Carborane Derivatives. An Eight-Membered be 1,4-dioxa-6,7-( 1,2-carboranylene)-2-( 1'-carboranyl)-2-methylExocycle of 1,2-Dicarbaclovododecaborane(l2) cyclooctane. T h e infrared spectrum contained absorptions a t 3060 cm.? for CH of carborane, a t 2560 cm.-l for BH of carBY NATHAN MAYESAND JOSEPH GREEN borane, a t 1374 cm.-l for CCH3, a t 1127 cm.-' for COC, and a t 730 for BH deformation of carborane. No absorption for Received February 15, 1965 OH was present. The molecular weight obtained cryoscopically in benzene was 378 (calculated 386). Anal. Calcd. for CgHSo02The syntheses of five- and seven-membered carBto: C, 27.96; H, 7.82; B, 55.93. Found: C,26.77; H, 7.92; borane' exocycles, in which one carborane participates B, 55.72. T h e methanol-water mother liquor was evaporated t o yield a in the ring, have been We wish to report gum from which unreacted 1,2-bis(hydroxymethyl)carborane now such an exocycle which has eight members. This was extracted with 57, aqueous sodium hydroxide and unreacted compound, 1,4-dioxa-6,7-(lj2-carborany1ene)-2-(1'-car1-(epoxyisopropeny1)carborane was extracted with petroleum boranyl)-2-methylcyclooctane, is a condensation prodether. T h e remaining solid, weight 9.6 g. (19% yield), was uct of 1-(epoxyisopropyl)carborane4 and 1,2-bisidentified as P-hydroxy-P-( carboran-1-y1)proppl-( 1 '-hydroxymethylcarboran-2'-yl)1nethyl ether. After three recrystallizations (hydro~ymethy1)carborane.~ The condensation yielded, from heptane it melted a t 196-201". T h e infrared spectrum in addition to the cyclooctane, the linear ether, exhibited absorption bands a t 3370 cm.-l for OH stretching and P-hydroxy-P- (carboran- 1-yl)propyl- (1'- hydroxya doublet a t 1170 and 1160 cm.-l for tertiary OH deformation. methylcarboran-2'-yl)methyl ether. The linear ether Other bands were a t 3080 cm.-l for CH of carborane, a t 2940 is probably an intermediate of the cyclooctane ; howand 2860 cm.-l for aliphatic CH, a t 2560 cm.-' for BH of carborane, a t 1116 cm.-l for COC, and a t 727 an.-' for BH deforever, this was not proven by independently converting mation of carborane. T h e molecular weight determined crythe linear compound to the cyclic compound. oscopically in benzene was 415 (calculated 405). T h e proton The preparation and probable structures of these magnetic resonance spectrum, obtained in benzene solution with compounds are illustrated in the equation below. tetramethylsilane added as an internal reference, exhibited the following singlets ( T ) : 8.95, CH3; 7.18, CHZ; 6.89, tertiary OH; CH? 6.72, CH2; and 6.26, -CBloH&H. The CHzOH group apBFJ / HOCHzC-CCH20H + CHz-C peared as an AB2 multiplet v i t h the center of the CHz doublet \O/ \o/ \ a t T 6.56 and the center of the OH triplet a t T 8.01. The firstC-CH BioHio order coupling constant was 7.9 C.P.S. The error for all values, \O/ except the -CBloHloCH shift, is &0.01 p.p.m. T h e -CBloHioCH BioHio shift error is greater due t o its unusually broad resonance. Anal. Calcd. for CgHsZOsBzo: C, 26.71; H, 7.97; B, 53.44; OH, 8.40. Found: C, 26.95; H 8.09; B, 51.34; OH, 8.92.
".
+
/
\
CHz \ '0
\
CHz
\
o,
CHq-C-C-
CH
Experimental 1-(Epoxyisopropeny1)carborane (26.0 g., 0.130 mole), 1,2bis(hydroxymethy1)carborane (26.0 g., 0.127 mole), and 150 ml. of toluene were placed in a reaction flask fitted with a water condenser, a Dean-Stark distilling receiver, a Teflon-coated magnetic stirrer, and a gas inlet tube. The mixture was heated with stirring t o reflux and became homogeneous. Boron trifluoride was bubbled into the solution several times during 48 hr. of refluxing, and 2.0 ml. of water was collected in the Dean(1) The term "carborane" is used here to denote 1,2-dicarbaclovododecaborane(l2). (2) S. Papetti, B. B. Schaefer, H. J. Troscianiec, and T. L. Heying, Inovg. Chem., 3, 1444 (1964). (3) J. Green, N. Mayes, A. P. Kotkoby, and M. S. Cohen, J . Polymer Sci., A2, 3135 (1964). (4) M. M. Fein, J. Bobinski, N. Mayes, N. Schwartz, and M. S. Cohen, Inovg. Chem., 2, 1111 (1903). ( 5 ) M. M. Fein, D. Grafstein, J. E. Paustian, J. Bobinski, B. M. Lichstein, N. Mayes, N. S. Schwartz, and M. S . Cohen, ibid., 2, 1115 ( 1 9 6 3 ) .
Acknowledgment-We wish to thank Xessrs. Larry Adlum and Raymond Storey for infrared analysis and Dr. David Kates for n.m.r. analysis. This work was supported by the U. S . Air Force, Edwards Air Force Base, under contract AF 33(616)-5639.
CONTRIBUTION FROM IXORGASIC MATERIALS DIVISION, NATIONAL BUREAU OF STASDARDS, WASHINGTON 25, D. C., A N D DEPARTMENT OF CHEMISTRY, UNIPERSITY O F MARYLAXD, COLLEGE PARK, &IARYI,ANn
Additional Observations on t h e Electronic S p e c t r u m of Copper(I1) Acetate M o n o h y d r a t e BY CURTW. R E I M A N NGERALD ,'~ F. KOKOSZKA,'~ AND GILBERT GORDOW'~
Received February 16, 1965
Copper(I1) acetate monohydrate is a binuclear complex.2 At low temperatures it is diamagnetic and (1) (a) National Bureau of Standards; (b) University of Maryland. (2) J. N. van Niekerk and F. R. L. Schoening, Acto Cryst., 6, 227 (1053).
NOTES 1083
Vol. 4 , No. 7, July 1965
20
I
I
I
I
I
I
I
I
I
I
19
18
17
16
15
14
13
12
II
IO
WAVENUMBER x
Figure 1.-Unpolarized single crystal spectrum of copper acetate a t 77°K. T h e peak a t about 14,700 cm.-l has been instrumentally flattened relative to the 11,000 crn. --I peak.
a t higher temperatures it is p a r a m a g n e t i ~ . ~The electronic spectrum of this most interesting material has been investigated. 4,6 The polarization properties of two absorption bands, one a t 14,400 cm.-l and the other a t 27,000 cm.-', have been detern~ined.~Our interest in re-investigating this spectrum arose when survey spectra revealed the presence of a previously unreported band in single crystals a t approximately 11,000 cm.-'. The new absorption band appeared as a shoulder on the 14,400 cm.-l band a t room temperature but became quite pronounced a t liquid nitrogen temperature (Figure 1). Graddona has also observed a band a t about 11,000 cm.-l in solutions containing dimeric copper(I1) alkanoates. The crystal spectra were taken on a Cary Model 14 spectrophotometer. Crystals were ground by hand to a thickness of approximately 0.2 mm., and the thickness was further reduced by etching with water. Spectroscopic measurements with polarized light were carried out on these crystals in the regions 800011,000 and 15,500-30,000 cm.-I. Since the polarization properties of the 14,400 cm.-l band have been reported4,5 and because of the large extinction coefficient in this region, this band was not studied in detail. Measurements in the regions indicated above showed that the apparent band center' of the 11,000-14,400 cm.-l bands was sensitive to the direction of polarization of incident light. This indicates that the 11,000 and 14,400 cm.-l bands do not have the same polarization properties, since the apparent band center can change with the direction of incident polarized light only if the polarization directions of the 11,000 cm.-l band differ from that of the 14,400 cm.-l band. Measurements on the [110] face indicated that the apparent band center occurred closest to the 11,000 cm.-l band (3) B. Bleaney and K. D . Bowers, Phil. Mag., 43, 372 (1952); H. Kumagai, H. Abe, and J. Shimada, Phys. Rev., 87, 385 (1952). (4) S. Yamada, S. Nakamura, and R. Tsuchida, Bull. Chem. SOC.Japan, 30, 953 (1957). (5) M. L. Tonuet, S. Yamada, and I. G. Ross, Trans. Faraday SOC.,60, 840 (1964). (6) D. P. Graddon, J . Inorg. Nucl. Chem., 14, 161 (1960); 17, 222 (1961). (7) We define the apparent band center as the bisector of lines which connect points of equal extinction coefficient in an extinction coefficient vs. energy plot where measurements are made in the wings of a baud or band systems.
when the light was polarized along the direction in which the z-polarized 27,000 cm.-l band was maximized. Measurements on crystals polished perpendicular to the b axis revealed that the apparent band center again occurred closest to the 11,000 cm.-l band when the z-polarized 27,000 cm.-l band reached a maximum. Measurements on other crystal faces showed the same correlation between shift in apparent band center toward 11,000 cm.-l and the intensity of the 27,000 cm.-' band. Thus, we have concluded that the 11,000cm.-' band is z-polarized. With the assignment of the 11,000 cm.-l band as z-polarized i t is noted that copper(I1) acetate monohydrate has three absorption bands in the 8000-30,000 cm. -l region with the indicated polarizations : 11,000 cm.-' ( z ) , 14,400 crn.-l (x, y), and 27,000 cm.-l (2). Recent magnetic measurements* indicate that the individual Cu(I1) ions in dimeric copper(I1) acetate monohydrate are only weakly coupled and that the energy levels can be described in terms of a single, six-coordinated cupric ion and the ground state in terms of the "hole" formalism can be described as essentially dXLy2. With this choice of ground state, the polarizations of symmetry-allowed transitions based upon three possible choices of effective symmetry a t a cupric ion can be summarized as is shown in Table I. TABLE I PREDICTED POLARIZATIONS FOR OPTICALTRANSITIONS FOR
-
Transition x2
- yz + 2 2
x2 - y2 x2
x2
xy
- y2 + xz
- yz
'-f
yz
COPPERACETATE C4"
CZ"
... ...
c2
z
z
XY XY
*..
z
X
x,Y
Y
z, Y
The observed spectrum can now be considered in terms of the requirements of these symmetry groups. 14,400 Crn.-l Band (xy-Polarized) .--This band can be assigned to the d,2-yz--* d,,, d,, transitions since these transitions are symmetry-allowed along x and y in C4" and CZand along x and y, respectively, in (2%". The shifts in peak position of this band with direction of polarization of incident light,5 however, and the existence of two in-plane g values suggest a choice of symmetry lower than CA~,. 27,000 Cm.-l Band (z-Polarized).-The origin of this band has not been established but if i t is a d-d band it can be assigned as the dZ2-y2 +- d,, transition, where presumably the stabilization of the d,, level of one cupric ion has occurred through the axial perturbation by the other cupric ion in the dimer. The dX2-y9 +d,, transition is symmetry-allowed along z in both groups, CzVand CP. Under the group C4", the vibronic mechanism must be invoked in order to account for the dX2-y2 --* dS2 transition. 11,000 Cm. -l Band (z-Polarized) .-The dXLyz --t d,, transition is symmetry-allowed along the z axis (8) G. Kokoszka, H. C. Allen, Jr., and G. Gordon, J . Chem. Phys., in press.
Inorgunic Chernislry
1084 NOTES in Czv and Cz and vibronically-allowed along z in Cb,.. The dXLy2 + d, transition is allowed along z in CZ symmetry and can only occur through vibronic coupling in Czv and C4v. Thus, if the dxLy24 d,, transition accounts for the 27,000 cm.-l band, the 11,000 em.-’ band can be assigned as the d22-y24 d,, transition. If the 27,000 cm.-l band (which does seem high for a copper d-d transition) does not correspond to dz2--y2 -+ d,,, then the 11,000 cm.-l band may correspond to either or both the d,2-y24 d,, and the d,,_,? 4 d,, transitions. Hansen and Ballhausen9 have recently reviewed the theoretical aspects of this problem and have proposed a new interpretation of the electronic spectrum of copper(I1) acetate monohydrate. References to previous theoretical work may be found in their paper. (9) A. E. Hansen and C. J. Ballhausen, to be published
Figure 1.-BloH1212. The boron and hydrogen atoms are numbered conventionally and each iodine atom takes the number of its nearest neighbor boron atom. The twofold (2) axis is perpendicular to the line B-1-B-3.
ponents and Coulomb integrals one wishes and allows several choices for resonance integrals. We report here results for two alternate choices of the resonance CONTRlBUTION FROM BOEINGSCIENTIFIC RESEARCH integral: Hij = KS,, with a value of -18.0 e.v. for LABORATORIES, SEATTLE, WASHIXGTON I< and Hi, = 0.5K’(Hii H j j ) S i jwith a value7 of 1.75 for K’. The program also computes the populaAn Extended Hiickel Theory tion matrix and charge distribution in the molecule. To obtain the positions of the atoms in BloHl2I2 Investigation of the Electronic Structure of we synthesized the BloH& results of Schaeffer8 and 2,4-Diiododecaborane(12), BloH1212 the results of Moore, Dickerson, and Lipscomb.g Synthesis is necessary because Schaeffer’s X-ray work BY EMMETT B. MOORE, JR. located only the iodine atoms with respect to the boron Received February 19, 1965 framework and not the boron atoms individually in the BloHlzIz crystal. We forced the atomic positions We have carried out an extended Huckel theory into Czv symmetry because the molecules have nearly (EHT) molecular orbital (MO) investigation’ of the this symmetry in the crystal, even though the space electronic structure of B10H12I2 in order to predict some groups8P1O require only CZ symmetry. The structure of its physical properties, as has been done for B10H14 of B10H1212 is derived from B10H14 by replacing hydrogen in similar ~ t u d i e s , ~and - ~ specifically to determine atoms 2 and 4 by iodine atoms (Figure 1). theoretically the polarizations of the first allowed elecFor the hydrogen Is, boron 2s and 2p, and iodine 5p tronic transitions in order to compare them with the Coulomb integrals we chose Mulliken’s valuesll of the experimental polarizations observed by Smallwood valence state ionization energies, which are - 13.60, and Eberhardk5 For this purpose we have used -15.36, -8.63, and -11.16 e.v., respectively. For Hoffmann’s IBM 7094 E H T p r ~ g r a m . ~This , ~ prothe iodine 5s Coulomb integral we chose -21.09 e.v., gram computes molecular orbital coefficients and enerwhich is consistent with Mulliken’s other values. For gies for up to 68 atomic orbitals given the atomic posithe hydrogen and boron Slater exponents we used 1.0 tions of up to 20 hydrogen atoms and up to 17 atoms of and 1.3, respectively. Since the presently available principal quantum number 2. The basis atomic E H T program does not handle atoms of principal orbitals are 1s Slater orbitals for hydrogen and 2s, quantum number 5, w e varied the iodine Slater ex2p,, 2p,, and 2p, Slater orbitals for the other atoms. ponent and found that 1.3 gave values for the overlap The program computes all ss, s u ) uu, and m r type overintegrals involving iodine orbitals which are good lap integrals and resolves them into appropriate inapproximations to, but somewhat less than, the correct tegrals between basis orbitals. The program also values. allows one to use whatever values for the Slater ex-
+
(1) For a more complete discussion, see E. B. Moore, Jr., “The Electronic Structure of BioHinIz,” Boeing Scientific Research Laboratories Document D1-82-0352, March 1964. (2) E. B. Moore, Jr., L. L. Lohr, Jr., and m7. L7. Lipscomb, J . Chem. Phys., 35, 1329 (1961). (3) E. B.Moore, Jr., ibid., 37, 675 (1962). (4) R . Hoffmann and W. N. Lipscomb, ibid., 37, 2872 (1962). (5) E. Smallwood and W. H. Eberhardt, private communication. (6) R. Hoffmann’sextended Huckel theory computer program is available from the Quantum Chemistry Program Exchange, Chemistry Department, Indiana University, Bloomington, Ind.
Results Our results are summarized in Table I for both BioH14 and B1oH12Iz. Both molecules have stable, closed(7) R. Hoffmann, J . Chem. Phys., 39, 1397 (1963). (8) R. Schaeffer,J . A m . Chem. Soc., 79, 2726 (1967). (9) E. B. Moore, Jr., R. E. Dickerson, and TI‘. S . Lipscomb, J . Chem. P ~ Y S a7, . , 209 (1957). (10) J. S. Kasper, C. M , Lucht, and D. Harker, Acta Cvyst., 3, 436 (1950). (11) R. S. Mulliken, J . Chem. Phys., 2, 782 (1934).
